2.0 FSI Poor idle when cold

[MEL744]

Haven't checked the purge valve, but I find it fascinating that the fuel level could even play a role in this issue. So now you have changed the purge valve what happens when the fuel level is low? Does it run rough when fuel is below a quarter of a tank?

I had a quick first go at opening the maps on WinOLS today - seems like it will be a full on job to work whats going on there but will persist. Today the fuel level was around 1 notch below 1/4 and the 1200 idle for 80 seconds returned. It wasn't 100% smooth, felt like there was a few misses and even a couple more after it settled to 700 - it was noticeably smoother than normal though. My current hypothesis that at least some of the rough running is due to carbon build up throughout the intake and this is probably what I was noticing, and the rest (majority) is due to some other problem.

Given the fuel level is low and the car seems to be operating in this alternate mode I'm keen to get as much data out of it before I need to fill it up again. Current list of things that could be worth checking: battery voltage as mentioned earlier, timing data and misfire count as a means of quantifying the 'smoothness'. Any other suggestions?

Made a few logs over the past few days, haven't plotted any as for these I don't think it will be particularly illuminating - most values are constant.

First up, a cold start with full fuel
- Rough idle, 1200 rpm observed for approx 40 seconds
- 56 misfires
- Battery voltage around 14.2V

Next, cold start with low fuel
- Smoother idle, 1200 rpm observed for approx 80 seconds
- Measured temp values as requested
- Temp values all seem reasonable

Finally, hot start (car was sitting for 2 hours or so)
- Same temp values as above

From these, the temperature adaptation factor discussed earlier seems to be the same for both the hot and cold starts - so I think its just some function of ambient air temp. Also interesting is that G299 (Intake air temp sensor 2 - the one in the air box) doesn't seem to show up anywhere, might have a look for it tomorrow in case it appears in some strange measuring block. I'm thinking the 1200 rpm idle could be triggered when the difference between the 'Engine Start Temperature' (seems to be the same as Engine Outlet Temp) or some other temperature and the ambient air are below a certain threshold (close to 0). This would represent a situation where the car has been sitting sufficiently long that it has been able to reach (close to) thermal equilibrium with its surroundings and so there for will require some special procedures to bring the emissions systems up to operating temp. Time also seems to play a factor, given how if you restart it, it will settle to 800 rpm immediately. Really need the maps to confirm though.

[saabman]

You certainly get a lot more misfires than I ever noticed - mine was down around 3 or 4.

I think the misfire data is the best metric for engine smoothness as I understand It is calculated by the variation in time between ignition points.

I’m at a loss as to what to check for which is why I was doing block map scan before and after staring the engine to try and capture everything and then look for differences - unfortunately I didn’t keep track of good and bad starts to properly correlate the data.

I didn’t notice any difference between bad purge valve, no valve and good valve. Apart from the engine not reporting gas fault once the new valve went in. I’ll check with nat andcsee I she has noticed any difference between full and empty. When I had the car here for a month I only used half a tank - I don’t work far from home ( benefit of rural city living)

Logging temp values might be give some clues as to when the high idle occurs - I don’t think it should happen as often as it does.

I’m going to organize to swap cars again so I can have a play and investigate further but I’ll order some injector seals so I can get right into the valves to clean them up. Does your engine have the tumble flaps? I need to remove the injectors to get the flaps out to get access to the valves. Might try and get some crushed walnut shells and blast the valve stems to clean them up.

[MEL744]

The difference in misfire count is slightly concerning... Will try doing some block scans. Also today it threw another CEL again for the EGR so really need to check that out. In the morning it was low on fuel but idled for only 40 seconds and was rough, so I have no idea what triggers the 80 second smooth mode. Pretty sure it has the tumble flaps, not 100% sure though (haven't checked myself).

[saabman]

I think we could be looking multiple faults. I had a lok through my logs last night and could find any with misfires logged, I never bothered to log them because when ever i did look at the misfire figgures they were so low.

Maybe concentrating on the missfires might be show up something. While the TB indicates that the 1200 rpm idle is normal Im not convinced it should happen as often as it does, in you case I think the missfires is a more significant issue and my be caused by some other fault.

Maybe log misfires for all cylinders at idle, high RPM, hot cold, etc. and see if it is confined to one cylinder or effects all of them.

Heres a couple of relevant Tech Bulletins Dropbox - TPI exhaust.pdf - Simplify your life
Dropbox - TPI Missfire.pdf - Simplify your life

[MEL744]

Logged the misfires for each cylinder today on a cold start as normal. Cylinder 1 recorded 73 misfires and 3 had 1. I have replaced the spark plugs and coil packs, so will need to look into this more. Don't have the time at the moment to be taking the intake manifold off to try swapping injectors.

[saabman]

Hmm definatly looks like a problem with #1 in some fashion. Might be time for a physical inspection! It only takes a few minutes to pull the manifold off but half a day to put those damn clips back on!

I found all my valves were pretty evenly coated in crud but not badly ( in my opinion anyway)

I’ve been working on trying to find out what the parameters are for the cold start high idle. I have been looking for a copy of the med9.5.1 funksionrahmen (that’s not spelt right) but with no luck - the best I’ve been able to find is for the 9.1 Im going to take stab in the dark and hope they are similar enough to get an idea as to what’s happening. Ive started translating the index using google translate. Once I’ve got the index finished I’ll post it up. I hope to then have an idea which sections are possibly relevant, I’ll then translate those which will be slow going. I might start another thread with the translation work.

I’ve also been contemplating the 500ohm reading for the heater resistance, that seems way to high - that equates to less than 300mW that’s the power of 2 LED dash lights - like nothing. You can generate more heat rubbing your hands together.

[saabman]

I tried taking a pic of the status of my valves - very difficult the flaps get in the way, in the blx engine injectors need to come out to remove the flaps

[MEL744]

Yep, will definitely try and inspect it after exams are done. Will be ordering the correct pliers for those hose clamps before starting this time! That is an interesting point you make about the resistance. I wonder if it is something else (not the resistance of the heating element). The exact labelling in VCDS is Resistance Bank 1 Sensor 1 and while it appears in the same measuring blocks as the heating data it isn't specifically stated as such. It's also worth nothing both B1 and B2 of S2 do not report resistances, so maybe the resistance reported is acting like an NTC thermistor on the wideband sensors? It should be possible to check the actual resistance of the heating element with a multimeter (the connectors are nice and easy to get to) but the problem is knowing which pins are which.

It's hard to say if that carbon build up is bad or not without having experience, have seen lots of images just like it on various forums but whether it is a problem or not is another story.

[saabman]

Ye sthe information leaves a lot to the imagination - Ive just finished translating the index 22 Pages it comes to

Ill post it up tomorow..

[saabman]

Heres a link to the index of the Med9.1 Funktionsrahmen Dropbox - MED9.1 Funktionsrahmen Index English.docx - Simplify your life

it is for a TSI engine but its the closest Ive found If anyone finds something better - ie a non turbo one or idealy the MED9.5.1 PLEASE let me have it and Ill re do it... this what Im working with Dropbox - 5-4420-01-41w038-pq35-0-frm-de-p-vag-med9-1-tfsi-zip - Simplify your life if anyone else wants to have a go at translating bits we can put it all together

The index still needs a bit of tidying up but its a start.
Ive also started on 4526 BGKSTDTA Cold start detection. Ive got a long way to go but here is a dump of part of it from google translate, It WILL change in coming days but I think starts to give some ideas. Dropbox - 4526 BGKSTDTA Cold Start Detection.docx - Simplify your life